1. society
2. crime

the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
1
Managing the Risk of Tagbacks to Leniency Applicants in
Cartel Investigations
Mich ae l G. Eg g e and Ale xandra L . Sh an d ell
T
The leniency program of the U.S. Department of Justice’s Antitrust Division and those modeled
after it by other jurisdictions provide incentives in the form of both carrots and sticks to compa-
nies to disclose cartel conduct. The “leniency-plus” program provides incentives and rewards for
companies already under investigation for involvement in one cartel to self-report participation in
different cartel conduct, i.e., to disclose new cartels to the enforcement agency. Yet where a
leniency-plus applicant reports a conspiracy involving new product areas (whether or not close-
ly related to those covered by the original leniency application), the original leniency applicant
risks being “tagged back” by the leniency-plus applicant. This risk threatens the balance of interests between the original leniency applicant and the enforcement agency, and has the potential
to disrupt an otherwise mutually beneficial relationship, particularly where the leniency-plus applicant reports a conspiracy involving a product or product area closely related to the original
leniency application.
Overview of Leniency and Leniency-Plus Programs
United States. In the United States, the Antitrust Division of the Department of Justice offers, under
certain conditions, complete amnesty from criminal conviction and fines to the first company to
identify itself as having participated in a criminal antitrust conspiracy. To obtain amnesty, a conspirator must be the first to report the conduct and must cooperate in the DOJ’s investigation and
prosecution of the other members of the conspiracy. The applicant must provide the DOJ with sufficient information about the industry and the conduct to demonstrate (1) that a criminal antitrust
violation has occurred, and (2) that the conspiracy is not one for which there has been a previous
amnesty applicant. The applicant must also show that it has taken all necessary steps to ensure
that it has promptly and completely ceased participation in the cartel.1
The DOJ provides the first applicant to come forward with a marker,2 protecting its status as
“first in” and allowing the applicant a period of time in which to “perfect” the marker. To do so, the
applicant must further elaborate on the criminal conduct reported to ensure that it meets the
䡵
requirements of the DOJ’s Corporate Leniency Policy and demonstrate to the DOJ its willingness
Michael G. Egge is a
to cooperate. During this period, the applicant has the opportunity to conduct an internal investi-
partner and Alexandra
gation to identify the full extent of its involvement in the conspiracy. The applicant will then prof-
L. Shandell is an
associate in the
Washington, D.C.,
1
See U.S. Dep’t of Justice, Corporate Leniency Policy (Aug. 10, 1993), available at http://www.justice.gov/atr/public/guidelines/0091.pdf.
2
A marker, memorialized in a letter issued by the DOJ, serves to block later applicants from seeking to obtain leniency for conduct that the
office of Latham &
initial applicant has already disclosed to the DOJ. The marker holds the first applicant’s place while it gathers evidence to support its initial
Watkins LLP.
representations to the DOJ regarding the conspiracy.
the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
2
fer evidence to the DOJ to define the “scope” of this conduct, for which the DOJ will grant the
applicant conditional amnesty. The DOJ usually defines the scope of the conduct simply by the
products or the industry coordinated by the cartel. A conditional amnesty letter then memorializes
the relevant scope of conduct and assures applicants amnesty from prosecution for the extent of
the described conduct, assuming the applicant’s full compliance with its cooperation obligations.
The successful applicant eliminates its criminal exposure, including corporate criminal fines,
entirely. Should the applicant fail to satisfy its cooperation obligations, the conditional amnesty
may be revoked. The DOJ states that only one out of over 100 conditional amnesty letters has ever
been revoked.3
In 1999, six years after implementing the present iteration of the amnesty program, the DOJ’s
Antitrust Division introduced the first leniency-plus program (referred to in the United States as the
“amnesty-plus” program).4 The leniency-plus concept is fairly straightforward and simple: where
a firm that has been identified as a participant in a cartel for which the amnesty position is no
longer available (i.e., when someone else has already ratted the firm out), that firm can secure a
significant discount from the criminal fine the DOJ recommends to the court for participation in the
original cartel if it is the first to provide information to the DOJ on a new cartel outside of the scope
of the first-reported conspiracy.5 It is called “leniency plus” because the leniency-plus applicant,
if successful, gets full immunity from sanction in the second-reported conspiracy as well as a
recommendation of a discount on penalties for its participation in the first-reported conspiracy—
the “plus.” The discount on the fines arising from the first-reported conspiracy will be greater than
if the leniency-plus applicant had only cooperated in the investigation of the first-reported conspiracy.6 This discount, as with any general discount for cooperation, is applied as a recommended reduction to the applicable sentencing guidelines for the conduct.
On the flip side, companies under investigation for participation in a cartel that do not disclose
their participation in separate cartels can be subject to “penalty plus” once those other cartels are
uncovered. Then, where a company fails to disclose additional anticompetitive conduct, depending on the circumstances, the DOJ will recommend a much higher starting point for the sentencing guidelines range than the company would otherwise receive.7
3
See Scott D. Hammond & Belinda A. Barnett, U.S. Dep’t of Justice, Frequently Asked Questions Regarding the Antitrust Division’s Leniency
Program and Model Leniency Letters 25 (Nov. 19, 2008), http://www.justice.gov/atr/public/criminal/239583.pdf. The amnesty revocation
cited here is likely a reference to the Stolt-Nielsen case. Stolt-Nielsen S.A. was a successful applicant for conditional amnesty for its
participation in a cartel in the parcel tanker industry. See Press Release, Stolt-Nielsen S.A., Stolt-Nielsen Transportation Group Granted
Conditional Amnesty in Parcel Tanker and Inland Barge Investigations (Feb. 25, 2003), http://www.stolt-nielsen.com/Media-Centre/FeedNews.aspx?link=http://cws.huginonline.com/S/154/PR/200302/893240.xml. Sometime after its receipt of the conditional amnesty letter,
the DOJ suspended Stolt-Nielsen’s conditional amnesty, alleging that the company had failed to promptly end the illegal conduct in which
it had engaged and to cooperate fully with the investigation. See Stolt-Nielsen S.A. v. United States, 352 F. Supp. 2d 553, 568 (E.D. Pa. 2005).
Ultimately, a federal grand jury returned an indictment against Stolt-Nielsen. United States v. Stolt-Nielsen S.A., 524 F. Supp. 2d 609, 610
(E.D. Pa. 2007). The district court granted Stolt-Nielsen’s motion to dismiss the indictment, id. at 627–28, and the DOJ declined to appeal.
See Interview with Scott Hammond on Stolt-Nielsen, G LOBAL C OMPETITION R EV. (May 2008), available at http://www.justice.gov/atr/
public/speeches/234840.pdf.
4
The U.S. leniency-plus program is called “amnesty plus” in the United States because of the unique features of the U.S. cartel enforcement
regime, most importantly, that cartel conduct is enforced criminally. Given that most other cartel enforcement regimes refer to amnesty plus
as “leniency plus,” we use “leniency plus” to refer to the U.S. amnesty plus program.
5
Hammond & Barnett, supra note 3, at 8–9.
6
Id.
7
Scott D. Hammond, Measuring the Value of Second-In Cooperation in Corporate Plea Negotiations, Remarks Before the 54th Annual ABA
Section of Antitrust Law Spring Meeting 7 (Mar. 29, 2006), available at http://www.justice.gov/atr/public/speeches/215514.pdf.
the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
3
Tagbacks in the Leniency Regime. When a company elects to take advantage of the leniencyplus program, it necessarily must identify its co-conspirators in the second cartel for which it seeks
leniency. One of the co-conspirators identified in the second cartel could sometimes be the
leniency applicant in the first cartel. This situation is a tagback because the identification of cartel participants in the first application—the “tag”—has set off a chain of events, informed by powerful incentives, resulting in the first cooperator being “tagged back” as a co-conspirator in the
second cartel.8
There are several varieties of tagback scenarios, some more serious than others. Tagbacks
may be aptly categorized in much the same way as burns to the human skin: in degrees. Likely
[T]he identification of
the most serious tagback scenario—one creating first degree burns—is where the leniency applicant is clearly the first mover in reporting wrongdoing in a given industry anywhere in the world.
cartel participants in
That applicant delivers to the DOJ a conspiracy as to a particular product in an industry where
there are a range of closely related, but distinct, products offered by suppliers. As is often the case
the first application—
in this setting, cartel conduct may not be limited to the one reported product so the investigation
of the first cartel leads, whether aided by leniency applications or not, to discovery of multiple car-
the “tag”—has set off
tels covering different product markets.
In short, the first mover produces the proverbial “whale” to a regulator: a conspiracy that allows
a chain of events,
the regulator to discover and prosecute a series of separate conspiracies involving related cartel
conduct across the industry. The first degree burn results from the first mover’s getting tagged
informed by powerful
back, through leniency plus granted to another cartel participant, early in the game, and perhaps
more than once. In this scenario, the first mover does not enjoy to the fullest the advantages of
incentives, resulting in
having been first in, even though the first mover is arguably responsible for the discovery of every
cartel identified thereafter. Second and third degree burns would be tagbacks materializing
the first cooperator
among parties removed from the first mover, with differences in degree turning on how far they are
removed. First-degree tagbacks raise the most acute potential for leniency policy disruption.9
Adoption of Similar Leniency Concepts in Other Jurisdictions. After the DOJ implemented its
being “tagged back”
revised amnesty program in 1993, several jurisdictions observed its success and implemented
as a co-conspirator in
similar leniency programs modeled on the DOJ’s approach. These leniency programs at times
vary in their approaches. All, however, follow the philosophy that to bust cartels, agencies need
the second cartel.
to give participants incentives to come forward by providing a sufficiently transparent and predictable system that includes (1) guaranteed immunity from penalty and (2) severe penalties for
non-disclosure or non-cooperation.
8
Notably, the hypothetical example provided by the DOJ in its Leniency Program FAQs does not involve a tagback. In other words, the initial
amnesty applicant was not a participant in the second cartel. Hammond & Barnett, supra note 3, at 9. Indeed, it is unclear whether a tagback has occurred in the fashion we describe in the United States. Yet it is undeniable a situation like that could arise, given the balance of
interests at play.
9
There are also different types of tagbacks. A cross-jurisdictional tagback would be where the first mover only moves first in one jurisdiction but fails to think about or act quickly enough in other jurisdictions. Once the first mover has triggered an investigation in one jurisdiction, the other cartel participants secure first-in leniency status elsewhere without having to count on leniency-plus opportunities. It is a
stretch to ever rate this scenario as a first degree burn, at least from a fairness perspective, given the evolution of global cartel enforcement
and defense practices. In short, the tagback in this instance is the first mover’s own fault. Another type, at least in theory, is a temporal tagback, where the first mover only secures the full immunity position for a specific period of time and a co-conspirator tags the first mover
back for the same product but a different period of time. Finally, there is the “different customer” tagback where the original leniency position was defined by price fixing or bid rigging as to a particular customer only, and a co-conspirator secures leniency-plus protection for
related conduct as to other customers for the same products.
the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
4
For example, the European Commission of the European Union, which issued its new Leniency
Notice in 200210 and revised it in 2006,11 offers complete immunity from fines for participation in
an illegal cartel to the first company to report the conduct and provide evidence of the cartel. As
in the United States, the EC issues the first applicant a marker during which that company may
conduct an internal investigation and proffer evidence to achieve a grant of immunity. The applicant must fully cooperate and immediately take steps to end its participation in the cartel.12
Unlike the United States, however, leniency applicants to the EC who are not first in may still
qualify for a guaranteed reduction in fine if they cooperate and provide “added value” to the investigation. The Leniency Notice specifically states that the first to come forward after the immunity
winner will be entitled to a 30 to 50 percent discount on its fines; the second, a 20 to 30 percent
reduction; and the third, a reduction up to 20 percent.13 (This is distinguishable from the DOJ’s
assertion that second-in cooperators will receive a discount on their criminal penalties. Although
the DOJ’s policy is to recommend a reduction in the sentencing guidelines, the second-in company must still plead guilty to the conduct and its penalty will be subject to the sentencing judge’s
discretion.)
While the EC’s 2006 Leniency Notice does not expressly contemplate either leniency-plus or
penalty-plus programs,14 it is possible that reporting and thereby qualifying for full immunity
regarding a different cartel could help secure the maximum discount available to a second- or
third-in leniency applicant in the original reported cartel.15
Canada’s immunity and leniency programs operate in much the same way as the U.S. amnesty
program. Canada’s Competition Bureau offers immunity from prosecution to the first to disclose
anticompetitive conduct.16 Companies and individuals who are not the first to report can still
receive a recommendation of leniency from the CCB—in the form of a discount—if they cooperate in the investigation. The CCB offers “immunity plus,” to later applicants who are eligible for
leniency for a particular offense if they disclose evidence of an additional anticompetitive criminal offense. Under the immunity-plus program, the immunity-plus applicant must meet the requirements of the immunity program for the new conduct. The immunity-plus applicant will then be eligible for immunity for the new conduct, plus an additional discount for its cooperation in the
original offense—typically an additional 5 to 10 percent.17
10
Commission Notice on Immunity from Fines and Reduction of Fines in Cartel Cases, 2002 O.J. (C 45) 3, available at http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:C:2002:045:0003:0005:EN:PDF.
11
Commission Notice on Immunity from Fines and Reduction of Fines in Cartel Cases, 2006 O.J. (C 298) 17, available at http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:C:2006:298:0017:0022:EN:PDF.
12
Id. at 18.
13
Id. at 20.
14
See G LOBAL L EGAL G ROUP LTD ., T HE I NTERNATIONAL C OMPARATIVE L EGAL G UIDE
15
EU courts have established a doctrine of non-discrimination, under which parties in comparable situations must be treated equally within
TO :
C ARTELS & L ENIENCY 2011 at 82 (2011).
a case. See Case T-311/94, BPB de Endraacht v. Comm’n, 1998 E.C.R. II-01129. This doctrine may limit the availability to the enforcement
agency of expressly granting more lenient treatment to a company in one matter for its cooperation in another, so as to avoid linking cases
together and potentially violating the non-discrimination principle.
16
Can. Competition Bureau, Bulletin—Immunity Program under the Competition Act § C (June 7, 2010), available at http://www.competition
bureau.gc.ca/eic/site/cb-bc.nsf/vwapj/Immunity-Program-2010.pdf/$FILE/Immunity-Program-2010.pdf.
17
Can. Competition Bureau, Bulletin—Leniency Program § 3 (Sept. 29, 2010), available at http://www.competitionbureau.gc.ca/eic/site/cbbc.nsf/vwapj/LeniencyProgram-sept-2010-e.pdf/$FILE/LeniencyProgram-sept-2010-e.pdf.
the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
5
In Canada, a tagback is, at least in theory, considerably more costly than in the United States.
To qualify for immunity in Canada, the applicant must disclose “any and all” competition offenses in which it may have been involved.18 By definition, a tagback means the first immunity applicant failed to disclose anticompetitive conduct in which it was involved. Thus the first immunity
applicant risks not only higher penalties for that second cartel, but also the loss of immunity for
the original conduct. For this reason, the CCB asserts it need not implement a “penalty-plus” program like the United States applies.19
Brazil, too, has implemented a leniency program to reward the first participant in a cartel to
[Leniency programs]
come forward to the Secretariat of Economic Law of the Ministry of Justice (SDE). Under the
leniency program, successful applicants are granted immunity from administrative sanctions and
are (or should be)
fines as well as criminal prosecution. As in the other jurisdictions described, the SDE requires that
the applicant be the first to come forward to disclose information about the anticompetitive con-
grounded in
duct and to cease its participation in the cartel.20 The SDE also offers leniency-plus benefits to
applicants who do not qualify for leniency in the prosecution of the first cartel disclosed. When a
transparency and
leniency-plus applicant comes forward with information about a second cartel that otherwise
meets the requirements of the leniency program, the company is entitled to immunity for the sec-
predictability, providing
ond cartel as well as a one-third reduction in its fines for the first cartel.21
the cartel participant
Leniency Policy Incentives and Tagback Risk
At the American Bar Association’s Antitrust Law 2012 Spring Meeting, Scott Hammond, Deputy
certainty that it will
Assistant Attorney General for Criminal Enforcement at the Antitrust Division, acknowledged that
tagbacks, although rare, present a real risk for amnesty applicants. Hammond stated that he had
receive the carrot of full
observed only one tagback in the last ten years, but was quick to validate that firms have good
reason to be concerned and to protect themselves.22
Applicant Incentives. Leniency programs are specifically designed to maximize incentives to
criminal immunity in
self-report and cooperate by pitting co-conspirators against one another in a race for protection
exchange for simply
against massive fines and, at least in the United States, reduced exposure to what can be enterprise-threatening private damage liability.23 They are (or should be) grounded in transparency and
cooperating and telling
predictability, providing the cartel participant certainty that it will receive the carrot of full criminal
immunity in exchange for simply cooperating and telling the enforcer what it knows. The alterna-
the enforcer what
tives, of course, are harsh fines and ruinous private damage exposure, if caught.
Enforcement agencies expect that, rather than face this risk of sanction and continue their con-
it knows.
duct, cartel participants will cease their illegal activity and self-report in exchange for a free pass
avoiding sanction. The risk of discovery and punishment is enhanced by the possibility that other
18
19
Can. Competition Bureau, supra note 16, § C.17.b.
Can. Competition Bureau, Adjustments to the Immunity Program § 5, available at http://www.competitionbureau.gc.ca/eic/site/cbbc.nsf/vwapj/Adjustments_to_the_Immunity_Program-e.pdf/$FILE/Adjustments_to_the_Immunity_Program-e.pdf.
20
Secretariat of Economic Law, Antitrust Division, Brazil Ministry of Justice, Fighting Cartels: Brazil’s Leniency Program 17, 19–21 (3d ed.
21
Id. at 24; see also art. 35-B, ¶ 8, Law No. 8.884, June 11, 1994, as amended by Law No. 10,149, Dec. 21, 2000 (Braz.).
2009), available at http://www.oecd.org/dataoecd/52/22/43619651.pdf.
22
Scott D. Hammond, Deputy Ass’t Att’y Gen., Antitrust Div., U.S. Dep’t of Justice, Remarks at ABA Section of Antitrust Law Spring Meeting
Panel Discussion on the Intricacies and Perils of Global Leniency Applications (Mar. 29, 2012) (unpublished).
23
Leniency applicants in the United States that have fully met cooperation obligations may avoid joint and several liability or treble damages;
their civil damage exposure is limited to their share of single damages assuming they cooperate with the civil plaintiffs. See Antitrust Criminal
Penalty Enhancement and Reform Act of 2004, Pub. L. No. 108-237, § 213(a)–(b), 118 Stat. 661, 666–67.
the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
6
co-conspirators may seize the opportunity and take the one and only pass available.24 This destabilizing effect can be profound as cartel participants struggle to meaningfully risk-adjust for the
likelihood that the cartel will be detected and punished. The DOJ’s relatively simple enhancements
to the transparency and predictability of its leniency policy in 1993 triggered a staggering increase
in applications and, in turn, guilty pleas for cartel conduct. Once accepted into the program, all
of the benefits are conditioned on full cooperation. The applicant has every incentive to carry
through with cooperation and protect its position.
Leniency plus generates even greater incentives to report new conspiracy conduct. First, once
the first mover has struck with an application in a given industry, the odds of the DOJ’s detecting
cartel conduct involving other products by any of the co-conspirators increase significantly even
without further leniency applications.25 The DOJ or any other skilled enforcer will have access to
documents and people and will do all it can to leverage the application to investigate conduct
related to other products whether in the same or separate industries.26 The heightened risk of
detection helps push the co-conspirator to consider leniency plus.27
Second, leniency plus offers greater incentives than the leniency program itself because of the
incremental discount the successful leniency-plus applicant receives off its fine for the first conspiracy, in addition to the full immunity it gets on the second conspiracy.28
Third, there is the prospect of simple, sweet revenge against the first mover. While revenge is
a rather base human motivation that is difficult to rationalize commercially or economically, regulators fully understand that humans run businesses. Leniency plus feeds nicely on the desire for
revenge.
Fourth, given the above three factors, the race for leniency over whatever cartel conduct exists
in the industry that is not already under investigation becomes incrementally more intense than if
there had been no leniency applications within the industry. The gloves are officially off.
Finally, the first mover’s application will trigger a highly invasive internal investigation at each
co-conspirator. For many firms, it is efficient to simply expand the investigation to the entire company and all product lines.
These enhanced incentives have returned remarkable results. The DOJ asserts that more than
half of its cartel investigations arise as leniency-plus disclosures resulting from investigations of
separate cartels.29
24
In addition to the incentives for a company to be first in, there are still benefits available to encourage a co-conspirator to cooperate as the
second, or later, in line. A later-in-time cooperator may be able to narrow the scope of the conduct for which it is prosecuted, earn a discount in its criminal fines for its conduct, and minimize the number of “carve-outs” for culpable executives from the company’s plea deal,
thereby ensuring that as few individuals as possible are subject to prosecution. Hammond, supra note 7, at 3–10.
25
The leniency application is not public, and does not normally become known until either the leniency applicant makes it public or it
becomes known later through civil litigation discovery or mere deduction from guilty pleas and other enforcement action. Once an application is made, however, it triggers or advances an investigation, which the DOJ eventually makes overt through the execution of warrants
and/or issuance of grand jury subpoenas.
26
For example, it will ask the infamous “omnibus question” of every witness from the applicant, and at every opportunity: whether the witness knows, under penalty of losing amnesty protection or obstruction of justice or both, of any cartel conduct in any industry and at any
time. Scott D. Hammond, Cornerstones of an Effective Leniency Program 15–16 (Nov. 22–23, 2004), available at http://www.justice.gov/
atr/public/speeches/206611.pdf.
27
As discussed later, co-conspirators in the earlier-reported conspiracy will generally discover that they have been tagged through one of two
ways: The DOJ may decide at some point to take the investigation public or the co-conspirator may itself come forward as a leniency applicant and learn that another company has already secured a marker for that conduct.
28
Hammond, supra note 3, at 9–10.
29
Hammond, supra note 26, at 9.
the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
7
Enforcer Incentives. The enforcement agency’s primary interest is to identify cartel conduct and
gather sufficient evidence to successfully prosecute as many cartels as efficiently as possible.
Investigating and prosecuting conspiracies is extremely difficult work. Leniency programs dramatically reduce that burden by not only identifying the existence of cartel conduct (in short, telling
the agency where to look) but also by producing invaluable admissions and eyewitness accounts
of conspiracy and other powerful evidence, without requiring the agency to lift a finger. The evidence is so powerful that the DOJ rarely has to try a cartel case. Co-conspirators routinely plead
guilty in the wake of a leniency application. Given these benefits and results, enforcers have a
powerful incentive to make leniency and leniency-plus as widely available as possible.
Nonetheless, even with a leniency applicant, the DOJ has to follow through on the investigation and prosecution, which is no easy task. The DOJ’s track record in trying cartel cases, even
with the benefit of leniency applicants and other cooperators, is not perfect.30 And while the DOJ
can leverage the leniency application process into requiring the applicant to conduct a full internal investigation and produce its fruit, it is ultimately the DOJ’s job to convert the leniency application into guilty pleas or convictions. The DOJ therefore has great interest in developing a strong
relationship with leniency applicants so that the DOJ can best harvest, understand, and assimilate the information it receives from each applicant. It generally does not want to do anything that
will alienate applicants or cool the pace and quality of cooperation.
Further, the DOJ has a strong incentive not to disrupt the transparency and predictability of the
leniency program. An applicant needs to know that it is going to be treated fairly and that by
coming in and cooperating, it will be given every opportunity to remain protected from prosecution.31 Any perceived unfair treatment by the DOJ will necessarily shake the applicant’s trust in
the program.
The Destabilizing Impact of Tagbacks. These incentives demonstrate a powerful alignment of
interests between leniency applicants and enforcers. Both want the relationship to be mutually
beneficial—the applicant wants to preserve its protection and avoid the DOJ’s being anything but
pleased with its cooperation (since a perceived failure to cooperate puts the applicant’s protection in jeopardy). The DOJ, for its part, wants the applicant to provide high quality, reliable information and wants to trust the applicant to thoroughly investigate any and all wrongful conduct. Any
30
See, e.g., United States v. Dutton, No. 3:02-cr-00220-RLV-1 (W.D.N.C. Jan. 12, 2004) (acquittal of NanYa employee Bradley Dutton in a case
where the DOJ had the benefit of an amnesty applicant and a second-in cooperator who pleaded guilty); United States v. Kim, No. 4:06-cr00692-PJH-3 (N.D. Cal. Mar. 21, 2008) (acquittal of DRAM employee Gary Swanson under similar circumstances). Following Dutton, the
DOJ subsequently shut down its investigation of the other two implicated firms in the four-firm polyester staple market.
Most recently, a federal jury returned split verdicts in AU Optronics, finding AU Optronics, its U.S. subsidiary, and two executives guilty
of conspiring to fix prices of liquid-crystal display panels (LCDs), but acquitting two more individuals, and resulting in a mistrial for failing
to reach a verdict with respect to one individual. See Press Release, Department of Justice, Taiwan-Based AU Optronics Corporation, Its
Houston-Based Subsidiary and Former Top Executives Convicted for Role in LCD Price-Fixing Conspiracy (Mar. 13, 2012), available at
http://www.justice.gov/atr/public/press_releases/2012/281032.pdf. Samsung Electronics Co., Ltd., and its subsidiaries obtained leniency for
their cooperation with the DOJ in providing information about the LCD cartel. Several other LCD suppliers subsequently pled guilty as well.
But AU Optronics went to trial, and, despite the cooperation of Samsung and others, the DOJ failed to obtain a guilty verdict for three of the
five individuals indicted. United States v. AU Optronics Corp., No. 3:09-cr-00110-SI (N.D. Cal. Mar. 13, 2012) (Document # 851, Special
Verdict Form, and Document #852, Transcript of Proceedings).
31
Indeed, the DOJ asserts that where an internal investigation reveals illegal conduct not covered by the scope of the conditional amnesty
letter, the original applicant is, in most cases, provided expanded coverage by an addendum to the letter so long as that entity meets the
criteria for amnesty on that conduct and continues to fully cooperate. See Scott D. Hammond, When Calculating the Costs and Benefits of
Applying for Amnesty, How Do You Put a Price Tag on an Individual’s Freedom?, Remarks at the Fifteenth Annual National Institute on
White Collar Crime at 4 (Mar. 8, 2001), available at http://www.justice.gov/atr/public/speeches/7647.pdf.
the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
8
disruption to this alignment threatens to doom the relationship and, in some circumstances,
potentially impair the transparency and predictability of the leniency program altogether.
The greater the threat of a tagback to a leniency applicant, the greater is the likelihood that the
balance of interests between the applicant and the agency will be disrupted. Take, for example,
a leniency applicant that has received a marker from the DOJ for a range of products and, in the
short time allotted to it under the marker, perfects the marker as to some, but not all, of the products in the range covered by the marker. Suppose also that there are other closely related products within the same industry, which are manufactured and sold by the applicant but are not covered by the marker, yet there are no cartel investigations involving this industry anywhere else in
the world. Under these circumstances, the DOJ would give the applicant a conditional amnesty
letter that covers the products for which the DOJ believes the applicant has perfected its marker.
The greater the threat
At this point, tagback risk to the leniency applicant is relatively low because the investigation has
yet to become public and no one else knows they have been tagged.32 Cartel participants learn
of a tagback to a
that they have been tagged either when they come forward to seek amnesty and find out that a
leniency applicant,
the investigation public.
co-conspirator has already secured a marker for the conduct or when the DOJ decides to make
The moment the DOJ investigation becomes public, however, tagback risks increase dramatthe greater is the
ically as targets quickly learn the amnesty position is taken and start to think about leniency-plus
opportunities. In this hypothetical, those opportunities are ripest for products not covered by the
likelihood that the
original marker but they also exist for the remaining products covered under the marker that were
not included in the conditional amnesty letter, at least to the extent that the DOJ does not extend
balance of interests
the marker letter.
between the applicant
by the first mover’s conditional amnesty letter. The DOJ may believe that, having already given at
This situation creates a dilemma for the DOJ because it wants to identify cartels not covered
least some opportunity to the original applicant to perfect the marker for the other products covand the agency will
ered by the marker and to conduct an internal investigation for all other related products, the
be disrupted.
uct lines. But if the DOJ allows a tagback in this situation, it could leave the initial applicant mate-
leniency-plus candidates may be in a position to provide better information on the related prodrially worse off through no fault of its own (apart from simply running out of time to perfect its marker for the other products or secure additional markers). The DOJ arguably would be responsible
for creating the tagback risk in the first place by going public in an investigation that would not
have happened but for the first mover’s application. An acceptance by the DOJ of a leniency-plus
position that allows a tagback under these circumstances, where the DOJ does not first give the
first-mover applicant the opportunity to seek protection for the products covered under the leniency-plus application, would be perceived as grossly unfair by the first-mover applicant. The firstmover applicant, feeling double-crossed, would surely fight any indictment on whatever grounds
possible (e.g., detrimental reliance). Moreover, that applicant may cease its cooperation. Most
importantly, however, the applicant—and all observers—will have lost confidence in the predictability of the system.
The DOJ can mitigate this risk by giving the first-mover applicant ample opportunity to protect
itself. This is done by communicating areas of potential concern to the first-mover applicant and
giving that applicant a chance to investigate and secure protection, either by marker or conditional
32
There does remain nonetheless a risk that any cartel associated with the products not covered by the marker provokes an independent leniency application by another firm.
the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
9
amnesty letter. Broad deference should be the rule for first-mover applicants lest the leniency-plus
program ultimately undermine the foundations of the leniency program—transparency and predictability. In short, as a policy matter, the leniency program should have priority over the leniency-plus program. The first-mover applicant can also mitigate tagback risk by doing everything in
its power to investigate and secure prompt protection, by marker or otherwise.
The DOJ appears to acknowledge the destabilizing effects tagbacks can have on the success
of the amnesty program by giving first-mover applicants wide berth to protect themselves against
tagbacks. In his 2012 ABA Antitrust Section Spring Meeting remarks, Hammond explained that to
enable applicants to avoid tagbacks and to fully incentivize information sharing, the DOJ does
attempt to grant broad amnesty to the initial applicant.33 Hammond acknowledged the temptation
presented to the DOJ when other companies or co-conspirators come forward offering additionA narrowly construed
al information that the initial applicant has not provided. He stated that the DOJ makes an effort
to avoid this temptation by offering the initial applicant the opportunity to investigate the same
marker leaves room
information. Hammond confirmed that the DOJ does not want to allow co-conspirators to tag back
first-in applicants.
Responsibility for avoiding tagbacks, however, does not rest solely on the DOJ. When the DOJ
for a co-conspirator
says it gives ample opportunity for the amnesty applicant to investigate whether it participated in
to enter into the
cartel conduct for other products revealed by an amnesty-plus request, it is also conveying that
surrounding space
amnesty protection.
and seek leniency for
Mitigating the Risk of Tagbacks
related conduct
take steps to ensure that it does not fall victim to the disclosures of a former co-conspirator.
the original amnesty applicant must do its part, investigate, and report to expand the scope of its
Given the risk that tagbacks pose to the original leniency applicant, the leniency applicant must
Secure as Broad a Marker as Possible from the Outset. A marker blocks a leniency-plus position,
through the leniency-
and hence a tagback. The applicant should seek a marker that will cover as much of the compa-
plus program.
reasonably amenable to extensions so long as the applicant demonstrates it is investigating and
ny’s product line as possible. Once a company has secured a marker, enforcement agencies are
being diligent about cooperating and perfecting its marker for all products. A narrowly construed
marker leaves room for a co-conspirator to enter into the surrounding space and seek leniency for
related conduct through the leniency-plus program.
Thus, in negotiating the marker language with the DOJ, the applicant should seek to use broad
and inclusive product definitions that capture any product markets that may have been touched
by collusive activity. In defining the product scope, the applicant should consider both the specific products at issue as well as the greater system of which those products and any related components may be a part. The applicant should seek an inclusive marker definition, rather than seek
a marker for specific products.
Understand the Clock. Leniency applicants have a timing advantage and should use it. As discussed above, tagback risks materialize once a co-conspirator learns of an investigation. That can
occur by the co-conspirator’s seeking a marker on its own and learning that one is not available,
or once the enforcement agency issues subpoenas, or the investigation is made public. An applicant cannot control when a co-conspirator presents itself for leniency on its own before the investigation becomes public, but it can understand when the agency is likely to go public with its
33
Hammond, supra note 22.
the antitrust source
䡵 www.antitrustsource.com
䡵 J u n e 2 0 12
10
investigation, and develop a plan for investigating related products not covered by the conditional
amnesty letter or the marker.
Be Prepared to Investigate Comprehensively. A leniency applicant needs to factor into its application decision the total life cycle costs of its application, and that must include expending considerable resources investigating products that are not covered within the scope of its protection.
Prioritize Internal Investigation Targets Based on Tagback Risk. A leniency applicant should be
able to figure out the sources of any tagback risk. Co-conspirators are known, so a leniency applicant should prioritize its investigation by focusing on business lines that overlap with co-conspirators (and that are not otherwise covered by the applicant’s leniency application) because that is
an obvious point of tagback exposure. The leniency applicant should carefully examine all product markets in which it operates and in which each of its co-conspirators operate. Tagback risk
exists in any market where there is overlap between the two companies involved in a cartel
because this is where the greatest incentive is for co-conspirators to hunt down collusive conduct
and report to the agency in an attempt to reap the rewards of the leniency-plus program. The risk
is thus greater where the companies at issue are diversified-product companies selling a variety
of related and non-related products.
Build a Trusting Relationship with the Enforcement Agency. As discussed above, if the applicant
proves to the agency that it will cooperate fully in providing all of the information requested of it,
the agency is more likely to maintain a constant dialogue and check with the applicant first if
another company comes forward with evidence of conduct in related products. Ideally, the applicant wants the enforcer to be mindful of tagback risks and look out for the applicant by asking if
it makes a given product and whether anything reported by the applicant is related to the product. This level of discourse should happen where the applicant has proven itself to the enforcement agency as a strong cooperator. In all events, where this inquiry occurs, the applicant should
assume there is a potential tagback risk and work with all deliberate speed to head it off.
Conclusion
Tagbacks fueled by the leniency-plus incentive structure pose real risk to leniency applicants and
enforcement authorities. Applicants must sufficiently protect themselves because incentives exist
for both former co-conspirators and enforcement agencies to pursue leniency-plus opportunities.
The applicant must factor the necessity of mitigating this risk into its analysis when applying for
leniency and determining its course of action. Enforcement agencies, for their part, should be cognizant of the destabilizing effects tagbacks can have on leniency programs and provide firstmover applicants in an industry ample space to protect themselves against tagbacks. 䢇